"Family Finder"

"Family Finder" - FTDNA's newest product offering - now in Beta Testing

Updated March 11, 2010

I am getting a lot of questions about FTDNA's "Family Finder"; so let me see if I can answer at least some of them.

First, let me say that my current experience comes from the similar 23andMe test - which appears to have been what stimulated FTDNA's newest product offering.  I got lucky and caught 23andMe's "Complete Edition" during a promotion.  (23andMe's "Complete Edition" includes the "Ancestry Edition" and the "Health Edition")  23andMe's  "Ancestry Edition" at $399 appears to be very similar to FTDNA's "Family Finder" for $249.  I was experimenting - so I tested 4 people at 23andMe - my Father, my Mother, my Wife and myself.

I placed orders on Feb 16 through the FTDNA Beta offering for my Wife and my Father - which will give me a direct comparison - and give me access to FTDNA's database that will quickly grow - probably much faster than the one at 23andMe.  (FTDNA doesn't have a DNA sample on hand for my Mother - so I couldn't do her test in the Beta - but I will order it when I can)

2/19/10 update - FTDNA advises:  "FTDNA’s “Family Finder” uses the latest microarray and platform from Affymetrix, while 23andMe uses the Illumina chip.  There are hundreds of thousands of SNPs that overlap between the two tests, but they do not totally coincide. It’s like comparing our Y-DNA test with one from another company. You will have matches between the two tests, but the match list and the relationships will not be as good and tight as when comparing 2 people that tested within FTDNA.  We have a much more solid algorithm to confirm relationships, which will present our results in a substantially more "genealogically friendly” way, as this is our constituency and we developed this product from the outset as a product for genealogists."

So - given FTDNA's 2/19 quote (above), I guess we'll have to wait and see how the 23andMe "Ancestry Edition" and FTDNA "Family Finder" tests compare.  My wife's results should be in the 1st or 2nd batch - and I am willing to share the outputs from both tests privately with one or two analytical folks who can analyse them to see where the genetic genealogy community will stand in dealing with the two sets of results.

One of my frustrations with the 23andMe test has been how clunky it is to use, with challenges in connecting to my matches and no tools to then help us figure out how we match.  I am also frustrated by the low percentage of responses from our 23ndMe matches, which is worsened by the large portion who don't know their ancestry. Overall (for my Mother, Dad and Wife) - about 40% have responded to my introductory contact and slightly less than 40% of those have provided ancestral info.  I've met a few nice folks, but have made only 2 actual connections so far.

I am expecting that FTDNA's experience in serving the genetic genealogy community will give us a much more powerful set of tools to find and understand our matches.  I can't wait to see how they do it!

General Comment: if both your Father and Mother are still living (and you can afford it) - test both of them - instead of yourself.  If both can't furnish you a DNA sample, then your own sample is almost as good.  I find that many of my matches are as good as my parents to a specific person, but that some matches are slightly weaker and a few matches are noticeably weaker.  This is a case where you do gain by testing the older generation - but you have to test two parents to be equivalent to yourself.

1. What type of test is this? The test constructs “autosomal blocks” using numerous SNPs in our autosomal DNA. The DNA you inherit from your parents is a mixed combination of the DNA they inherited from each of their parents. However, DNA that is close together on the chromosome tends to be inherited together. Sections of DNA that are passed on intact are called “autosomal blocks”. Your DNA consists of autosomal blocks inherited from many of your various ancestors. The more distant the ancestor, the smaller the blocks you tend to inherit from that ancestor, and the more possible it is that you will not inherit any DNA from that ancestor. By comparing the number and size of the autosomal blocks you inherit in common with other people, you can match from siblings and parents to 3rd and 4th cousins and, possibly to 5th cousins - or even further.  I am told by FTDNA that there will be no yDNA or mtDNA SNPs reported - so (unlike 23andMe) you won't get any insights on your y or mt haplogroup.  (note - I changed this last point after Raymond Wing and Steven Perkins both advised I had an error and FTDNA confirmed it)

As example - I am a 52.99% match to my Mother and a 47.67% match to my Father - which shows that the DNA evaluated by the test isn't a 50/50 blend of ancestry from my parents.  (The explanation is the inclusion of the X-Chromosome)   The 23andMe test confirmed my Deep Clade - which I already knew from Deep Clade testing at FTDNA. 

My two examples of success:

A. My wife found a 3rd cousin by comparing ancestral surnames with a match.  (He shares 0.51% on 4 blocks - a really nice match)   Both she and her newly-found cousin have Medley and McWilliams as ancestral surnames - and we quickly found that Billie Medley and Betsie McWilliams of Amory MS are their most recent shared ancestors

B. My Mother found a 4th cousin by comparing ancestral surnames.  (He shares 0.21% on one block - a rather weak match)   In this case, both have the surname "Hassell" in their ancestral surnames.  I went to my very complete listing of the Hassell  family and quickly found that there wasn't a "Mary V Hassell" as his wife reported  - but did spot a "Mary E Hassell" who seemed promising.  After Mother's new cousin's wife actually checked her information (instead of working from memory) - she confirmed that the name was "Mary Ellen Hassell" - with details which fit perfectly into my Mother's family.  Mary Ellen Hassell's husband and family info was "missing" from our Hassell compilation.  The two share most recent common ancestors of Zach Hassell and Phebe Parker.  In summary -  I found a missing branch of Mother's family and Mother's newly-found cousin discovered that he was descended from one of the most famous of the early Texas families - Parker.  (the family of the Fort Parker massacre, Cynthia Ann, Quanah, Texas Rangers and the founders of the first protestant church in Texas) 

2. What will the results tell me? FTDNA states that we'll be told our matches; their name and email address, and be able to download our raw data

We'll be told a suggested degree of relationship (probably something like "4th Cousin") and a probable range of relationship (like "3rd to 6th cousin") as well as the amount of DNA (which would be something like 0.41%) and number of autosomal blocks we share with our matches.   Like the yDNA testing, this is working with probabilities, not an absolute structure - so you won't know "exactly" how distant your cousin is from you.  Of course, once you make the connection, you will know exactly what the relationship is.

Recent discussions with FTDNA indicate that the useful level of the Family Finder test in comparing two individuals is "5th Cousins" - which is 7 generations - and better than the "5 Generations" that I had originally understood. 

3. What do the results look like? Like our yDNA and mtDNA tests, the raw "Results" are a mind-numbing array of information - in this case, SNP mutations.  Frankly, I haven't even looked at all 4 sets of my family's results.

What I work with are the reports of "Xth Cousin", the "Percentage DNA Shared", and the number of "Shared Blocks".  I use this info to target which person to contact and share ancestral surnames.

4. How do the results fit into my Project? In the early stage, it may not be clear, as there won't be many results.  This testing will work like the other tests - it's not what your data says - but who you match - and how do you match.

Some of the researchers are already comparing the results output and finding the common DNA among multiple people who match, so we should be able to partially reconstruct ancestral profiles over time.  (I haven't done this yet)  However - you need multiple matches to do this.  I expect that this process will happen fairly quickly and be particularly relevant for surname DNA projects trying to make sense of their unconnected branches in a Genetic family (Lineage).

FTDNA will have a new class of project available for families of a common ancestor who are testing specifically to research that common ancestor - or for groups who find themselves with a common ancestor.

5. Are results displayed along side Y-DNA results and mtDNA results? I don't know, but I doubt it.  We'll have to wait and see - but I am expecting this to be a separate section of information until we find a shorthand way to report it alonside of yDNA and mtDNA

6. Is there a template for these results? Again, I don't know - but our Genetic Genealogy community is imaginative and sharing - so the good ideas will emerge pretty quickly.

7. Will there be a new tab on our project pages? FTDNA says "Yes".   I expect that FTDNA will promote this extensively, so it should be highly visible in their ordering, public and project pages.

I expect that World Families will be adding another project page to our set of pages - but we'll need to see what we have to work with before we can create a page

8. Are the new results used with the older Y-DNA data and mtDNA data? The simple answer has to be "YES"!  The results will be reported separately, but new DNA information should help us figure out more specifically how closely folks who "match" are related.

As example - my Dad and a very sweet lady are perfect mtDNA FGS matches.  However, we can't find our paper connection and have theorized that an unrecorded adoption in the mid-1800s is the solution.  However, autosomal blocks shows no matching autosomal DNA.  So - either this new DNA tool isn't always helpful - or their most recent shared ancestor lived long enough ago to keep them from having any shared autosomal DNA - which argues against the "unrecorded adoption" solution.

9. Are there new calculators, charts, etc? Probably - we'll have to see what is available once we start getting results.  We know from experience that FTDNA will add and improve their tools and presentation as they gain experience.

In summary, FTDNA are in Beta testing.  The product is not available to the general population and I am not sure who is eligible for beta testing.  Many of these questions should be answered by the time FTDNA opens the ordering to all.

I discussed Family Finder in London with Bennett and Max and with the group of ISOGG members who were at the "Who Do You Think You Are?" show and have had another discussion with FTDNA staff since our return.  I am even more excited about the possibilities after hearing more of the plans and explanations.   I am expecting the database at FTDNA to be populated by folks who know their ancestry and who are interested in sharing. FTDNA's significantly better service and web tools, along with their lower price should cause FTDNA to quickly overtake 23andMe in database size - which is a huge issue. 

Bottom Line - autosomal block testing should be a great tool for the individual researcher to individually test and then randomly find their cousins who have independently chosen to test.  Possibly even more important will be the "focused" testing - by presumed cousins trying to confirm a weak or non-existent paper trail.  And - most exciting to me - will be the surname project based explorations.  For these, we'll test male and female descendants of the families who are matching on the Surname yDNA test to see if we can find the actual autosomal blocks inherited from a most recent common ancestor.  If we can (as I suspect), we should be able to reconstruct the Lineage's family tree and finally have a dna solution to make up for the lost paper trails. 

Here are a couple of useful links:



Please share your learnings and experiences.  I'll add to this as we move forward


Note - I tweaked this blog posting on 2/18/10 to improve the accuracy of my statements - based on input provided by FTDNA.